Everyone is excited (well almost everyone) about pipes and they should be. Mark looks at it form the web services point of view and sees it as a showcase for doing things the web way.
Dare has a critical view on it and puts his finger right on the weakness: in its current form, it's tricky to scale (if possible at all - right now it's slow as hell).
I agree with both. Pipes show the potential and really will be an incubator for things to come. But a central pipe service will not scale, nor does the internet community want to rely on such a thing. But yahoo! will drive the development here. I am sure google will soon release beta its own version. And Microsoft will make an anouncement that its version will be the next Vista Ultimate extra as part of its Live Brand or so. Death to WS-* for sure.
So I believe that pipes will grow and prosper, also on server as yahoo's, but mostly on each and every computer. I wrote a good two months ago on something similar to pipes which I called the funky web. It was not about pipes, but describes the processing model of pipes. My main point was that these things are able to run just anywhere.
If I were on yahoo's pipe team, I would push for the development of a screensaver that could run pipes. Process so and so many pipes and you get some yahoo goodies, a pro flickr account or such. Seti showed how powerful that approach is.
Pipes are great: they will run just anywhere on the web. They just need to PUT their results on some server somewhere, and keep the output URL stable.
Technorati Tags: pipes, webservices, yahoo